How did you come to be mediator of the Intergovernmental Authority on Development peace process?

I had been Kenyan special envoy in 1997-98; at the end of October 2001, President Moi called me and said 'I want to give you a job and I don't want you to refuse it.' I wanted to negotiate myself out of it, but he was determined. The international community had refused to support the peace process without a new mediator.

I went to the parties first of all. The Sudan People's Liberation Movement (SPLM) and President al-Bashir agreed to accept me as mediator. I found that the IGAD Secretariat had built a bill of 10 million Kenyan shillings for rent, services and allowances. My first task was to resolve this, firstly by getting the Kenyan government to pay for it (Moi directed his Treasury to do this) and secondly by establishing credibility with the donors. I went to the US, the UK and Italy, who all supported me.

How did you get the support of the international community for your initiative?

The communiqué of the Khartoum IGAD Summit in January 2002 called on the Chairman of the Committee on Sudan to 'rejuvenate the IGAD Peace Process and invite other initiatives with a view to coordinate the efforts.' I translated what that phrase meant: bringing on board anybody who had something to offer. I saw a chance to bring in the international community in the form of the IGAD Partners Forum. I gave myself powers to invite anybody according to my requirements, as long as they could pay for it. It included the British, the Norwegians, the Americans, the Italians and others.

But there was also at that time an Egyptian-Libyan Initiative (ELI) on Sudan, so I went to Cairo to assure the Egyptians that I was not going to do anything behind their back. President Mubarak was very good and said he would send his Minister in charge of security to go and talk to [Libyan leader] Gaddafi.

What was the role of the other IGAD member states?

The peace process was an IGAD initiative and therefore when I called the parties to meet in May 2002 I invited the other IGAD envoys, and it was attended by the Nairobi ambassadors of Ethiopia, Eritrea and Uganda. We formed a team and I also started looking for resource persons. In consultation with the envoys I asked for specialised people in particular areas, like Professor [Fink] Haysom from South Africa, a renowned constitutional lawyer, Professor Julian Hottinger and professors from the University of Nairobi.

What was the format of the negotiations?

In 1997-98 I had been following the IGAD Declaration of Principles (DoP) point by point, dealing with the issues of recognition, farming and so forth. The DoP was a complete diagnosis but somebody had to do the prescription for every problem. I realised this was not going to work.

My idea in 2002 was to have continuous negotiation on each element of the DoP until we were through, and prescribe there and then what would happen. When I called the parties in May 2002 they didn't have a problem with the programme of work or the modalities, but they could not agree on two words: the government insisted on a 'transition' period; the SPLM wanted an 'interim' period. They refused to sign the document.

The government wrote a stinking letter to me saying I was insolent and disrespectful to the government delegation and appeared to be a partial mediator. I didn't listen to them: they had wanted to continue but I had called the meeting off prematurely. I wrote a very nice letter back to them saying this had been one of the most successful meetings in the negotiations and invited them to come on 19 June to Machakos, Kenya. Surprisingly they all came. These were the first serious negotiations. In Karen, the delegations had been small, but in Machakos each delegation had eight negotiators plus a two-person secretariat.

The negotiations took a very interesting turn at Machakos. For 29 days the two sides were mostly shouting at each other. But while they were shouting I was making notes on the issues. After 29 days we prepared a text. President Carter had advised me that without a single negotiating text I risked losing the process. So I translated the DoP into a single text and then zeroed in on the two main issues: self-determination and the separation of state and religion. Wealth sharing, security, power sharing, the judiciary, civil rights and so forth were also in the framework.

On 19 July I decided that enough was enough, consulted with the observers and envoys, and we gave the text to the parties at about 10 pm. I added that I needed the answer that night! I gave them an hour, but they wanted two weeks. They started using their mobile phones to talk to their bosses. So the decisions were not being made in Machakos, but somewhere else. The teams had the power to negotiate, to stick to certain positions, but not to make final decisions.

By midnight they were not through. At one o'clock they called me and said, 'Yes, we have agreed on the two issues.' They said, 'Can we write them tomorrow?' I said 'No, you must write them tonight.' We wrote the framework protocol and both sides were happy about it, but would not sign. Salva Kiir, the signatory leader of the SPLM delegation was in Nairobi that night and so was Dr Ghazi of the government. So I prepared the document and the following day we signed it.

What was the role of the observers at Machakos?

The British, Norwegians, Americans and Italians were there. Their role was very positive, except each country had its own perception of what the resolution would be. In fact, when I made this single negotiating text the American observer asked if I had checked with Washington. I said I wasn't answerable to Washington. I actually didn't need to check with anybody, because I had been given the task and the mandate to negotiate a peace agreement. This was the case throughout, even when it looked like others wanted to take over the process.

How did you proceed after the Machakos Protocol?

We started talking again after a two-week break but in September 2002 the SPLM attacked Torit and the government pulled out of the negotiations. So I had to negotiate to get the parties to come back. Neither delegation was really keen to run away. I think there was a group who wanted to snub the negotiations, but there were many who recognised the negotiations were going somewhere.

We made the breakthrough with the agreement on a cessation of hostilities in October, which was one of the most satisfying bits of the process for me. I had said from the start: that there was no point in negotiating while fighting. But it wasn't going in until after Torit.

Then of course we had the Kenyan elections in December 2002. Moi stood down but that did not really affect my position in the peace process. I had a very able deputy in the army and I spent most of my time with the process.

About this time you started mediating on the contested areas of the Nuba Mountains, southern Blue Nile and Abyei outside the auspices of IGAD. Why?

When the SPLM negotiators went back to report the Machakos success, these 'conflict areas' were not in the framework, and they came back and insisted that this must be discussed. The government would not discuss the conflict areas. I proposed that we discuss them with the British or the Americans chairing. The SPLM said no, the British would side with the government. The government said no, the Americans would favour the SPLM. Eventually they agreed to negotiate these areas under the auspices of Kenya, but not IGAD.

I had to be very careful at this stage regarding my credibility with the envoys of the other IGAD countries. I said I would mediate but on the condition that I choose my resource persons and observers. The parties consented and I chose the IGAD envoys as observers and resource persons. There was a lot of symbolism. We wrote the agreement on the conflict areas on Kenyan-headed paper initially.

The main IGAD talks at this time were structured thematically, with groups focused on specific issues – was this because they were no longer coming together?

What I had sensed in Machakos was that these people did not have the right mandate. Secondly, all of them, the north and the south, were concerned about their constituencies and I wanted to get to the constituencies. I went to Sudan and met many people: civil society, religious groups, lawyers, judges, everybody. Having retired from the Kenyan army at the end of February 2003, I had time to travel. I went with the Machakos framework and tried to find out the positions of the parties regarding all the issues, the issues of security, power sharing, wealth sharing.

By June 2003 I had travelled enough in Sudan. I wrote the 'Nakuru Document.' The observers agreed that this was a document to resolve all the outstanding issues after Machakos. Of course, the government hit the roof. President al-Bashir went ballistic and said, "whoever wrote the Nakuru Document must soak it, drink it and go to hell!" He later sent an emissary to me say that this was just politics and he didn't mean it.

In August 2003 we went to Nanyuki in Kenya. The SPLM wanted direct negotiations, while the government wanted proximity talks. Unlike the SPLM, the government did not want to negotiate on the basis of the Nakuru Document. We got stuck for about two weeks, but I wouldn't agree publicly that we were actually stuck. I tried to pin down who was controlling the government side, and I succeeded. I wouldn't ever say how, but I found that they were reporting back to Sudanese First Vice-President Ali Osman Taha. And they were being told, 'Hold on, don't be the ones to pull out.'

I called the Kenyan Minister of Foreign Affairs, Stephen Musyoka, in Cairo and asked him to see President al-Bashir in Khartoum and ask for his vice-president to come and negotiate with Dr John Garang. Garang agreed to come to Naivasha on 1 September. Al-Bashir was reluctant, saying John Garang had twice snubbed his vice-president, and a third time would be really catastrophic. But eventually we agreed on a four-day meeting from 1 September – hoping that it was going to be very short.

For three days Garang didn't come! This was very difficult! His officers were insisting that he should only negotiate with the president himself. Two SPLM people helped me: Dr Justin Yaac Arop and Commander Deng Alor, who were in Nairobi. John Garang wrote a letter to Musyoka saying they would reschedule the meeting. Dr Justin gets the letter, puts it in his pocket, but doesn't give it to Musyoka. He puts pressure on Garang by writing to him that he should not dream of going back to Kenya, because the Kenyans were mad at him! They would not want to see him and his family would be kicked out.

On the third day at 6 o'clock John Garang arrives in Naivasha! By 6:30 we had put Ali Osman Taha and Garang together, the first time they had met face to face. They asked us to leave them alone to talk, to get to know each other.

I don't know what they said but Ali Osman later told me the decision rested on a bottle of water: there was only one bottle of water between them. If John Garang opened this water, then they were going to have an agreement. If it was left unopened there was not going to be an agreement. So John Garang opens the water, pours it for Ali and then for himself. For Ali the first test is over!

John Garang told me later that Ali intended to convince him to sign an agreement like the Khartoum Peace Agreement signed by Riek Machar in 1997. Of course he could not buy that. He would not have fought so long for an agreement like that.

It was important to have these two at Naivasha, but they also kept their delegations. They gave them things to go and discuss and come back and brief them. But every time I wanted to put something forward I'd go to see the two principals. If I wanted to talk one-on-one I'd always ask the rest to leave.

When we moved to the Simba Lodge at Naivasha, they asked, 'How long are you staying?' I said maybe a week. A week turned into 16 months until we signed the Comprehensive Peace Agreement! We were there full-time except for pre-agreed breaks like for Ramadan or breaks to go and consult. There were of course threats: sometimes the parties would pack their things, go to reception and say 'we are leaving!'

And you were there all the time, which is something not every mediator could have done.

Yes. I don't think anybody like a president would have the time to sit, eat, go to the gym, sauna or steam bath with the parties. Being involved with them in those different ways was very important. It takes months of engagement. It takes neutrality. I'm not one to close any avenues for anybody. And because I sometimes told off either party, I was perceived as neutral by both parties.

You were always negotiating with the international community looking over your shoulder. How difficult was that?

Very difficult. Especially the Americans, who were addressing local American politics. In early 2004 they even brought an aircraft and said, 'Let's sign an agreement half way so that the parties can attend the State of the Union Address.' I resisted that.

Sometimes it was positive. I brought in Colin Powell. I brought in Senator Danforth. They applied useful pressure. Whenever one party reneged, I always rang Colin Powell. He came to Nairobi to combat heel-dragging as we were trying to give the final push.

When we went to Naivasha the observers were no longer sitting with us. They came to consult with the parties and me about progress, but were more on the periphery than in the process. I had a rule that the swimming pool at the hotel was the last place the observers could come – the boundary. Beyond that it was only the parties, the envoys and members of the secretariat.

What issues were most difficult to reach agreement on?

Signing a security agreement establishing two armies in one country was probably the most difficult. I didn't believe it possible myself, but Garang told me he had already negotiated on this basis with Ali Osman Taha: you are trying to bury your head in the sand when you know that two armies exist so the best thing is for us to recognise these ones. There were more than two of course.

Second was the issue of wealth sharing. The government was concerned about its budget, but did not consider that this was a new dispensation – that it was not going to be business as usual. At one time we were negotiating over 2 1/2 per cent of the oil revenues and this became very difficult. There were not so many problems around power sharing. The parties negotiated it between themselves, but even here it was a matter of percentages and getting any movement was difficult. Eventually we found a formula.

What difference in negotiating skills and strengths did you notice between the two sides?

Both lacked capacities for mediation. I realised later that they hadn't really conceptualised a changed position – even the SPLM, who did not believe they were going to get an agreement. The government always negotiated on their own terms and had never imagined a new dispensation. But as they went on negotiating they started realising that they actually had to change. So it was really a process and to help it we brought in more resource persons, from the World Bank, from oil-producing countries, from Australia to talk on the issues of land. I brought in nearly everybody!

And that led in effect to the World Bank and the UNDP running their Joint Assessment Mission (JAM)?

Yes. This was after they realised there was going to be agreement. This was well integrated with our talks and I even had a person in the JAM. At one stage it was going faster than the negotiations, which was also good for other people to see and brought the realisation that there was going to be peace.

Is this an example of the 'holistic approach' to peace mediation you employed from mid-2003?

Yes. And of course this issue of involving everybody. Don't leave anybody out.

But involving everybody is difficult if the two sides want to be the only negotiators.

Of course. But what I'm saying is get everybody to buy into the ideas, not bring everybody to the table. After the Machakos Protocol everybody wanted to come in: the French, the Dutch, the South Africans, the Arab League. But the doors had to be closed. So I said that if the parties agreed then they would be accepted. But the two parties said, 'look, you are putting this problem on us and we really don't want to appear like we're refusing anybody. But what we will do is this: you forward whatever people apply and we will not reply.' So I would tell the would-be interveners, 'I'm still waiting for the parties.' And they would ask me which party so they could sort it out, and I would say 'both parties haven't replied.'

But also within Sudan there were other people who wanted to have their voice heard...

Oh yes, civil society. Everybody. And Ali Osman Taha and Garang also wanted to bring in as many people as possible after making any decision so they could buy the idea. They did it whenever they were signing an agreement: the generals for security arrangements, economists from both sides for the wealth-sharing agreement.

But why was it that the SPLM and the government of Sudan did not involve other groups?

I don't know. Initially in 1994 the SPLM-United and SPLM-Mainstream had agreed to negotiate from one side of the table. So I believe that was the main reason why the government negotiated with the SPLM, the most credible force at that time.

But of course the CPA itself is not comprehensive. Comprehensive in my understanding would be the whole of Sudan. That was never on the table: the government would not allow it. Every time I tried to raise it they said, 'oh, you want to come and resolve all our conflicts? Come to Darfur, come to Eastern Sudan, we have enough problems. Come to the north; we have a lot of problems!'

The agreement is still more comprehensive than any other agreement that has ever been made in Sudan, even the 1972 agreement, because it has an implementation modalities section that gives the who, where and when.

Looking back at the process now, what would you do differently?

I would not waste time on people who do not have the mandate. I'd get the parties to identify the issues, but then very quickly identify what they can and can't make decisions on and lift it to a level where the decisions are really made.

I would go earlier in the negotiations to the ground, to find out what the people want, which is not necessarily the same as what the negotiating team is presenting.

Finally, you have to tread very carefully. You have to be able to read the international community's interests way, way ahead. Because if you don't, you run the risk of colliding with them and losing support.

What would you recommend to international partners in any other peace process: how could they best support it?

Help the parties build their own capacity to negotiate credible agreements. The conceptualisation has to come from the parties themselves rather than from without. Early on the parties started wanting me to produce papers for both parties, so that they could attribute it to the Secretariat, claiming it wasn't their position. They were concerned about their people back home accusing them of selling out. But I learned this and I used to get the parties to write papers on issues. Only after they had given me their extreme positions would I try to bring them closer together. Each party would tell me what they would accept, but not in front of the other. I'd call two from each side and start broaching the paper, and each side would say, 'No, that is not our position!' And I'd say, 'I know, but these are your extreme positions and I have tried to consider each of you in drawing this position.' So they would say, 'OK, if this is the Secretariat's position, then we'll consider it.'

But you have to get them to have ownership of the agreement and you have to build their capacity. Even governments sometimes don't have the capacity. Or they send their blocking troops. You identify them and you don't even waste your time in trying to move their positions because they are only there to block progress.

Are there any other lessons for mediators?

I found it personally taxing to administer and mediate, but if you are mediating you must know how much money you have, for how long you can schedule the meeting – but then meetings never end on time. You have to have a big contingency and a good rapport and credibility with donors and account for their funds.

The team is important. In Somalia Bethuel Kiplagat had a big team and there was no way he could account for the money quickly with that many people. I was dealing with a small enough number of people to account for. I wasn't worried about money because we signed a budget agreement so that the donors gave money through GTZ of Germany and I accounted for it through GTZ. The funding was more or less open-ended, but of course no one expected the process to last three years, least of all Moi, who wanted it completed in a year!